In [ ]:
#importing the necessary libraries 
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from scipy.stats import binom,norm,poisson
In [ ]:
df=pd.read_csv('aerofit_treadmill.csv')
In [ ]:
df
Out[ ]:
Product Age Gender Education MaritalStatus Usage Fitness Income Miles
0 KP281 18 Male 14 Single 3 4 29562 112
1 KP281 19 Male 15 Single 2 3 31836 75
2 KP281 19 Female 14 Partnered 4 3 30699 66
3 KP281 19 Male 12 Single 3 3 32973 85
4 KP281 20 Male 13 Partnered 4 2 35247 47
... ... ... ... ... ... ... ... ... ...
175 KP781 40 Male 21 Single 6 5 83416 200
176 KP781 42 Male 18 Single 5 4 89641 200
177 KP781 45 Male 16 Single 5 5 90886 160
178 KP781 47 Male 18 Partnered 4 5 104581 120
179 KP781 48 Male 18 Partnered 4 5 95508 180

180 rows × 9 columns

In [ ]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 180 entries, 0 to 179
Data columns (total 9 columns):
 #   Column         Non-Null Count  Dtype 
---  ------         --------------  ----- 
 0   Product        180 non-null    object
 1   Age            180 non-null    int64 
 2   Gender         180 non-null    object
 3   Education      180 non-null    int64 
 4   MaritalStatus  180 non-null    object
 5   Usage          180 non-null    int64 
 6   Fitness        180 non-null    int64 
 7   Income         180 non-null    int64 
 8   Miles          180 non-null    int64 
dtypes: int64(6), object(3)
memory usage: 12.8+ KB
In [ ]:
df.isna().sum() 
# there are no values missing
Out[ ]:
Product          0
Age              0
Gender           0
Education        0
MaritalStatus    0
Usage            0
Fitness          0
Income           0
Miles            0
dtype: int64
In [ ]:
# to check in basic characteristics
df.describe()
Out[ ]:
Age Education Usage Fitness Income Miles
count 180.000000 180.000000 180.000000 180.000000 180.000000 180.000000
mean 28.788889 15.572222 3.455556 3.311111 53719.577778 103.194444
std 6.943498 1.617055 1.084797 0.958869 16506.684226 51.863605
min 18.000000 12.000000 2.000000 1.000000 29562.000000 21.000000
25% 24.000000 14.000000 3.000000 3.000000 44058.750000 66.000000
50% 26.000000 16.000000 3.000000 3.000000 50596.500000 94.000000
75% 33.000000 16.000000 4.000000 4.000000 58668.000000 114.750000
max 50.000000 21.000000 7.000000 5.000000 104581.000000 360.000000
In [ ]:
df['Income (in K)']=round(df['Income']/1000,2)
In [ ]:
# detecting the outliers
fig=plt.figure(figsize=(20,20))
k=1
for i in ['Age','Education','Usage','Fitness','Income','Miles']:
    plt.subplot(3,2,k)
    sns.boxplot(data=df,y=i,x='Product')
    plt.title(i)
    k=k+1
fig.suptitle('Checking the Outiers',fontsize=25)
plt.show()

From this data we observe that for column Education,usages, and fitness are consistent data with negligible outliers

Product Purchased¶

In [ ]:
# Number of products purchased 
Products=df['Product'].value_counts().reset_index()
Products
Out[ ]:
index Product
0 KP281 80
1 KP481 60
2 KP781 40
In [ ]:
sns.countplot(df['Product'])
plt.title('Product Purchased',fontsize=20)
plt.show()

d:\Anaconda\lib\site-packages\seaborn\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.
  warnings.warn(
In [ ]:
# calculating the Marginal Probability of the different products
Products['PMF']=Products['Product']/180
Products
Out[ ]:
index Product PMF
0 KP281 80 0.444444
1 KP481 60 0.333333
2 KP781 40 0.222222

A. Observation- Purchased order is KP281>KP481>KP781

B. Proability of Purchasing:

  • KP281=44.4%
  • KP481=33.3%
  • KP781=22.2%

Product Purchased by the Gender¶

In [ ]:
Product_Gender=pd.crosstab(df['Product'],df['Gender'],margins=True)
Product_Gender
Out[ ]:
Gender Female Male All
Product
KP281 40 40 80
KP481 29 31 60
KP781 7 33 40
All 76 104 180
In [ ]:
# plotting the countplot
plt.figure(figsize=(12,8))
sns.countplot(data=df,x='Product',hue='Gender')
plt.title('Product Purchased by Gender',fontsize=20)
Out[ ]:
Text(0.5, 1.0, 'Product Purchased by Gender')
In [ ]:
#Getting the conditional Probability 
Product_Gender['Product|Female']=round(Product_Gender['Female']*100/76,2)
Product_Gender['Product|Male']=round(Product_Gender['Male']*100/104,2)
Product_Gender
Out[ ]:
Gender Female Male All Product|Female Product|Male
Product
KP281 40 40 80 52.63 38.46
KP481 29 31 60 38.16 29.81
KP781 7 33 40 9.21 31.73
All 76 104 180 100.00 100.00
In [ ]:
#Calculating the Marginal Probability
Product_Gender['P(P^F)']=round(Product_Gender['Female']*100/180,2)
Product_Gender['P(P^M)']=round(Product_Gender['Male']*100/180,2)

Product_Gender
Out[ ]:
Gender Female Male All Product|Female Product|Male P(P^F) P(P^M)
Product
KP281 40 40 80 52.63 38.46 22.22 22.22
KP481 29 31 60 38.16 29.81 16.11 17.22
KP781 7 33 40 9.21 31.73 3.89 18.33
All 76 104 180 100.00 100.00 42.22 57.78

Probability - F=Female, M=Male

  • P(KP281|F)=40/76 = 52.63% ---- P(KP281|M)=40/104 = 38.46%

  • P(KP481|F)=29/76 = 38.16% ---- P(KP481|M)=31/104 = 29.81%

  • P(KP781|F)=07/76 = 09.21% ---- P(KP781|M)=33/104 = 31.73%

Observation :-

  1. KP281 is more purchased, But the chances with equal amount purchase:-
    • women Buying KP281(52.63%) >>> male buying KP281(38.46%)

  2. KP481 is more purchased with almost equal number of man and women but chances:-
    • women buying the KP481 (38.15%) >>> male Buying KP481(29.30%)

  3. KP781 is least Purchased
    • Predominately Purchased by the Males.
    • chances of Male buying the KP781 >>>> Female Buying KP781

  4. Chances of Buying KP781 in males are higher the Chances of Buying KP481 by males

Product Purchased on the basis of Marital Status¶

In [ ]:
Product_Marital=pd.crosstab(df['Product'],df['MaritalStatus'],margins=True)
Product_Marital
Out[ ]:
MaritalStatus Partnered Single All
Product
KP281 48 32 80
KP481 36 24 60
KP781 23 17 40
All 107 73 180
In [ ]:
# PLOTTING
plt.figure(figsize=(12,8))
sns.countplot(data=df,x='Product',hue='MaritalStatus')
plt.title('Product VS Marital Status', fontsize=20)
plt.show()
In [ ]:
# calculating the Conditional Probability 
Product_Marital['Product|Partnered']=round(Product_Marital['Partnered']*100/107,2)
Product_Marital['Product|Single']=round(Product_Marital['Single']*100/73,2)
Product_Marital['P(P^P)']=round(Product_Marital['Partnered']*100/180,2)
Product_Marital['P(P^S)']=round(Product_Marital['Single']*100/180,2)
Product_Marital
Out[ ]:
MaritalStatus Partnered Single All Product|Partnered Product|Single P(P^P) P(P^S)
Product
KP281 48 32 80 44.86 43.84 26.67 17.78
KP481 36 24 60 33.64 32.88 20.00 13.33
KP781 23 17 40 21.50 23.29 12.78 9.44
All 107 73 180 100.00 100.00 59.44 40.56

observation: KP281,KP481,KP781 is more Purchased by the Patterned customers

Probability - S=Single, P=Partnered

  • P(KP281|S) = 43.84% ---- P(KP281|P) = 44.86%

  • P(KP481|S) = 32.88% ---- P(KP481|P) = 33.64%

  • P(KP781|S) = 23.29% ---- P(KP781|P) = 21.50%

Observation :-

  • Irrespective of Partnered or single there are higher chances that they will buy KP281
  • All the Product have similar chances wrt to buying by partnered or single

Analyzing the Different Types purchased by age¶

In [ ]:
# Extracting age and product then Classification in age category
product_age=df[['Product','Age']]
product_age['Age_category']=pd.cut(product_age['Age'],bins=[0,20,26,30,35,40,45,50],labels=['<20','21-26','25-30','30-35','35-40','40-45','45-50'])

C:\Users\Ravi Ahuja\AppData\Local\Temp\ipykernel_684\1213120939.py:3: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  product_age['Age_category']=pd.cut(product_age['Age'],bins=[0,20,26,30,35,40,45,50],labels=['<20','21-26','25-30','30-35','35-40','40-45','45-50'])
In [ ]:
product_age
Out[ ]:
Product Age Age_category
0 KP281 18 <20
1 KP281 19 <20
2 KP281 19 <20
3 KP281 19 <20
4 KP281 20 <20
... ... ... ...
175 KP781 40 35-40
176 KP781 42 40-45
177 KP781 45 40-45
178 KP781 47 45-50
179 KP781 48 45-50

180 rows × 3 columns

In [ ]:
# Analyzing the Outliers of Age for each product
plt.figure(figsize=(12,10))
n=sns.boxplot(data=df,y=df['Age'],x='Product')
n.set_yticks(range(15,50,1))
plt.grid()
plt.show()

Outliers:-

  • KP281 - Age > 47
  • KP481 - Age > 45
  • KP781 - Age > 38 there are More outliers for KP781 Treadmill
In [ ]:
# we groupby the age and product then count the frequency and plotted a scatterplot
plt.figure(figsize=(12,10))
k=sns.scatterplot(data=df.groupby(['Product','Age']).size().reset_index().rename(columns={0:'values'}),
                  x='Age',y='values',hue='Product')
k.set_xticks(range(18,55,1))
k.set_yticks(range(0,15,1))
plt.ylabel('No of Purchase')
plt.show()
In [ ]:
Product_age_table=pd.crosstab(product_age['Age_category'],product_age['Product'],margins=True)
Product_age_table
Out[ ]:
Product KP281 KP481 KP781 All
Age_category
<20 6 4 0 10
21-26 35 27 19 81
25-30 14 4 11 29
30-35 11 17 4 32
35-40 8 6 2 16
40-45 3 1 2 6
45-50 3 1 2 6
All 80 60 40 180
In [ ]:
# calculating the Conditional Probability 
Product_age_table['KP281|Age']=round(Product_age_table['KP281']*100/Product_age_table['All'],2)
Product_age_table['KP481|Age']=round(Product_age_table['KP481']*100/Product_age_table['All'],2)
Product_age_table['KP781|Age']=round(Product_age_table['KP781']*100/Product_age_table['All'],2)
Product_age_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Age KP481|Age KP781|Age
Age_category
<20 6 4 0 10 60.00 40.00 0.00
21-26 35 27 19 81 43.21 33.33 23.46
25-30 14 4 11 29 48.28 13.79 37.93
30-35 11 17 4 32 34.38 53.12 12.50
35-40 8 6 2 16 50.00 37.50 12.50
40-45 3 1 2 6 50.00 16.67 33.33
45-50 3 1 2 6 50.00 16.67 33.33
All 80 60 40 180 44.44 33.33 22.22
In [ ]:
# plotting the histogram for each product WRT age
fig=plt.figure(figsize=(20,8))
k=1
for i in ['KP281','KP481','KP781']:
    plt.subplot(1,3,k)
    sns.histplot(df.loc[df['Product']==i]['Age'],bins=int(32))
    plt.title(i)
    k=k+1
fig.suptitle('Checking the Age for each Product Type',fontsize=25)
plt.show()
In [ ]:
# plotting the Age category 
fig=plt.figure(figsize=(20,8))
k=1
for i in ['KP281','KP481','KP781']:
      bin=[0,20,26,30,35,40,45,50]
      label=['<20','21-26','25-30','30-35','35-40','40-45','45-50']
      plt.subplot(1,3,k)
      sns.countplot(x=pd.cut(df.loc[df['Product']==i]['Age'],bins=bin,labels=label))
      plt.title(i)
      k=k+1
fig.suptitle('Checking the Age for each Product Type',fontsize=25)
plt.show()
In [ ]:
# Plotting the Count with age category 
plt.figure(figsize=(15,10))
sns.countplot(data=product_age,x='Age_category',hue='Product')
plt.title('Age VS Product Type',fontsize=20)
plt.show()
In [ ]:
Product_age_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Age KP481|Age KP781|Age
Age_category
<20 6 4 0 10 60.00 40.00 0.00
21-26 35 27 19 81 43.21 33.33 23.46
25-30 14 4 11 29 48.28 13.79 37.93
30-35 11 17 4 32 34.38 53.12 12.50
35-40 8 6 2 16 50.00 37.50 12.50
40-45 3 1 2 6 50.00 16.67 33.33
45-50 3 1 2 6 50.00 16.67 33.33
All 80 60 40 180 44.44 33.33 22.22

Observation-

  • KP281 - Customer having age in between 20-30
  • KP481 - Customer having age in between 30-35
  • KP781 - Customer having age in between >40(outliers),25-30

Product analysis with Income¶

In [ ]:
df
Out[ ]:
Product Age Gender Education MaritalStatus Usage Fitness Income Miles Income (in K)
0 KP281 18 Male 14 Single 3 4 29562 112 29.56
1 KP281 19 Male 15 Single 2 3 31836 75 31.84
2 KP281 19 Female 14 Partnered 4 3 30699 66 30.70
3 KP281 19 Male 12 Single 3 3 32973 85 32.97
4 KP281 20 Male 13 Partnered 4 2 35247 47 35.25
... ... ... ... ... ... ... ... ... ... ...
175 KP781 40 Male 21 Single 6 5 83416 200 83.42
176 KP781 42 Male 18 Single 5 4 89641 200 89.64
177 KP781 45 Male 16 Single 5 5 90886 160 90.89
178 KP781 47 Male 18 Partnered 4 5 104581 120 104.58
179 KP781 48 Male 18 Partnered 4 5 95508 180 95.51

180 rows × 10 columns

In [ ]:
#Extracting the Income and product column and defining the category income 
Product_income=df[['Product','Income (in K)']]
Product_income['Income_category']=pd.cut(Product_income['Income (in K)'],bins=[0,30,40,50,60,70,80,90,100,110],labels=['<30','30-40','40-50','50-60','60-70','70-80','80-90','90-100','100-110'])

C:\Users\Ravi Ahuja\AppData\Local\Temp\ipykernel_684\510611060.py:3: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  Product_income['Income_category']=pd.cut(Product_income['Income (in K)'],bins=[0,30,40,50,60,70,80,90,100,110],labels=['<30','30-40','40-50','50-60','60-70','70-80','80-90','90-100','100-110'])
In [ ]:
Product_income
Out[ ]:
Product Income (in K) Income_category
0 KP281 29.56 <30
1 KP281 31.84 30-40
2 KP281 30.70 30-40
3 KP281 32.97 30-40
4 KP281 35.25 30-40
... ... ... ...
175 KP781 83.42 80-90
176 KP781 89.64 80-90
177 KP781 90.89 90-100
178 KP781 104.58 100-110
179 KP781 95.51 90-100

180 rows × 3 columns

In [ ]:
#creating a table 
Product_income_table=pd.crosstab(Product_income['Income_category'],Product_income['Product'],margins=True)
Product_income_table
Out[ ]:
Product KP281 KP481 KP781 All
Income_category
<30 1 0 0 1
30-40 22 9 0 31
40-50 25 21 5 51
50-60 26 23 6 55
60-70 6 7 6 19
70-80 0 0 4 4
80-90 0 0 7 7
90-100 0 0 9 9
100-110 0 0 3 3
All 80 60 40 180
In [ ]:
# calculating the Conditional Probability 
Product_income_table['KP281|Income']=round(Product_income_table['KP281']*100/Product_income_table['All'],2)
Product_income_table['KP481|Income']=round(Product_income_table['KP481']*100/Product_income_table['All'],2)
Product_income_table['KP781|Income']=round(Product_income_table['KP781']*100/Product_income_table['All'],2)
Product_income_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Income KP481|Income KP781|Income
Income_category
<30 1 0 0 1 100.00 0.00 0.00
30-40 22 9 0 31 70.97 29.03 0.00
40-50 25 21 5 51 49.02 41.18 9.80
50-60 26 23 6 55 47.27 41.82 10.91
60-70 6 7 6 19 31.58 36.84 31.58
70-80 0 0 4 4 0.00 0.00 100.00
80-90 0 0 7 7 0.00 0.00 100.00
90-100 0 0 9 9 0.00 0.00 100.00
100-110 0 0 3 3 0.00 0.00 100.00
All 80 60 40 180 44.44 33.33 22.22
In [ ]:
# plotting the histogram for each product for Income
fig=plt.figure(figsize=(20,8))
k=1
for i in ['KP281','KP481','KP781']:
    plt.subplot(1,3,k)
    sns.histplot(df.loc[df['Product']==i]['Income'],bins=int(32))
    plt.title(i)
    k=k+1
fig.suptitle('Checking the Age for each Product Type',fontsize=25)
plt.show()
In [ ]:
# plotting the Income Category
fig=plt.figure(figsize=(20,8))
k=1
for i in ['KP281','KP481','KP781']:
      bin=[0,30,40,50,60,70,80,90,100,110]
      label=['<30','30-40','40-50','50-60','60-70','70-80','80-90','90-100','100-110']
      plt.subplot(1,3,k)
      sns.countplot(x=pd.cut(df.loc[df['Product']==i]['Income (in K)'],bins=bin,labels=label))
      plt.title(i)
      plt.xticks(rotation=340)
      k=k+1
fig.suptitle('Checking the Age for each Product Type',fontsize=25)
plt.show()
In [ ]:
# Plotting the income category with the product
plt.figure(figsize=(15,10))
sns.countplot(data=Product_income,x='Income_category',hue='Product')
plt.title('Income VS Product',fontsize=20)
plt.show()
In [ ]:
# Analyzing the Outliers of Income for each product
plt.figure(figsize=(12,10))
n=sns.boxplot(data=df,y=df['Income (in K)'],x='Product')
n.set_yticks(range(30,130,5))
plt.grid()
plt.show()

Outliers:-

  • KP281,KP781 there are no outliers
  • KP481- Outliers are above 60K
In [ ]:
# we groupby the age and product then count the frequency and plotted a scatterplot
plt.figure(figsize=(20,10))
k=sns.scatterplot(data=df.groupby(['Product','Income (in K)']).size().reset_index().rename(columns={0:'values'}),
                  x='Income (in K)',y='values',hue='Product')
k.set_xticks(range(30,130,5))
k.set_yticks(range(0,10,1))
plt.ylabel('No of Purchase')
plt.grid()
plt.show()
In [ ]:
Product_income_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Income KP481|Income KP781|Income
Income_category
<30 1 0 0 1 100.00 0.00 0.00
30-40 22 9 0 31 70.97 29.03 0.00
40-50 25 21 5 51 49.02 41.18 9.80
50-60 26 23 6 55 47.27 41.82 10.91
60-70 6 7 6 19 31.58 36.84 31.58
70-80 0 0 4 4 0.00 0.00 100.00
80-90 0 0 7 7 0.00 0.00 100.00
90-100 0 0 9 9 0.00 0.00 100.00
100-110 0 0 3 3 0.00 0.00 100.00
All 80 60 40 180 44.44 33.33 22.22

Observation-

  1. KP281- customer having Income in between 30k-60k
  2. KP481- customer having Income in between 40k-60k
  3. KP781- customer having Income in greater > 60k

Analyzation of Product with Education¶

In [ ]:
df
Out[ ]:
Product Age Gender Education MaritalStatus Usage Fitness Income Miles Income (in K)
0 KP281 18 Male 14 Single 3 4 29562 112 29.56
1 KP281 19 Male 15 Single 2 3 31836 75 31.84
2 KP281 19 Female 14 Partnered 4 3 30699 66 30.70
3 KP281 19 Male 12 Single 3 3 32973 85 32.97
4 KP281 20 Male 13 Partnered 4 2 35247 47 35.25
... ... ... ... ... ... ... ... ... ... ...
175 KP781 40 Male 21 Single 6 5 83416 200 83.42
176 KP781 42 Male 18 Single 5 4 89641 200 89.64
177 KP781 45 Male 16 Single 5 5 90886 160 90.89
178 KP781 47 Male 18 Partnered 4 5 104581 120 104.58
179 KP781 48 Male 18 Partnered 4 5 95508 180 95.51

180 rows × 10 columns

In [ ]:
# Import the sum of values with Education and Product
Prodcut_education_table=pd.crosstab(df['Education'],df['Product'],margins=True)
Prodcut_education_table
Out[ ]:
Product KP281 KP481 KP781 All
Education
12 2 1 0 3
13 3 2 0 5
14 30 23 2 55
15 4 1 0 5
16 39 31 15 85
18 2 2 19 23
20 0 0 1 1
21 0 0 3 3
All 80 60 40 180
In [ ]:
# Analyzing the Outliers of Education for each product
plt.figure(figsize=(12,10))
n=sns.violinplot(data=df,y=df['Education'],x='Product')
n.set_yticks(range(12,22,1))
plt.grid()
plt.show()
In [ ]:
# we groupby the Education in years and product then count the frequency and plotted a scatterplot
plt.figure(figsize=(20,10))
k=sns.scatterplot(data=df.groupby(['Product','Education']).size().reset_index().rename(columns={0:'values'}),
                  x='Education',y='values',hue='Product')
k.set_xticks(range(11,22))
k.set_yticks(range(0,50,2))
plt.ylabel('No of Purchase')
plt.grid()
plt.show()
In [ ]:
df['Education'].unique()
Out[ ]:
array([14, 15, 12, 13, 16, 18, 20, 21], dtype=int64)
In [ ]:
# calculating the Conditional Probability 
Prodcut_education_table['KP281|Education']=round(Prodcut_education_table['KP281']*100/Prodcut_education_table['All'],2)
Prodcut_education_table['KP481|Education']=round(Prodcut_education_table['KP481']*100/Prodcut_education_table['All'],2)
Prodcut_education_table['KP781|Education']=round(Prodcut_education_table['KP781']*100/Prodcut_education_table['All'],2)
Prodcut_education_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Education KP481|Education KP781|Education
Education
12 2 1 0 3 66.67 33.33 0.00
13 3 2 0 5 60.00 40.00 0.00
14 30 23 2 55 54.55 41.82 3.64
15 4 1 0 5 80.00 20.00 0.00
16 39 31 15 85 45.88 36.47 17.65
18 2 2 19 23 8.70 8.70 82.61
20 0 0 1 1 0.00 0.00 100.00
21 0 0 3 3 0.00 0.00 100.00
All 80 60 40 180 44.44 33.33 22.22
In [ ]:
# Plotting the countplot with Education vs product
plt.figure(figsize=(15,10))
sns.countplot(x=df['Education'],hue=df['Product'])
plt.title('Education VS Product',fontsize=20)
plt.show()
In [ ]:

Observation:-

  1. KP281= customer having Education of 14 and 16 years
  2. KP281= customer having Education of 14 and 16 years
  3. KP781= customer having Education of 18 and more years

Analyzation of Usages with the Prodcut type¶

In [ ]:
df
Out[ ]:
Product Age Gender Education MaritalStatus Usage Fitness Income Miles Income (in K)
0 KP281 18 Male 14 Single 3 4 29562 112 29.56
1 KP281 19 Male 15 Single 2 3 31836 75 31.84
2 KP281 19 Female 14 Partnered 4 3 30699 66 30.70
3 KP281 19 Male 12 Single 3 3 32973 85 32.97
4 KP281 20 Male 13 Partnered 4 2 35247 47 35.25
... ... ... ... ... ... ... ... ... ... ...
175 KP781 40 Male 21 Single 6 5 83416 200 83.42
176 KP781 42 Male 18 Single 5 4 89641 200 89.64
177 KP781 45 Male 16 Single 5 5 90886 160 90.89
178 KP781 47 Male 18 Partnered 4 5 104581 120 104.58
179 KP781 48 Male 18 Partnered 4 5 95508 180 95.51

180 rows × 10 columns

In [ ]:
Product_Usage_table=pd.crosstab(df['Usage'],df['Product'],margins=True)
Product_Usage_table
Out[ ]:
Product KP281 KP481 KP781 All
Usage
2 19 14 0 33
3 37 31 1 69
4 22 12 18 52
5 2 3 12 17
6 0 0 7 7
7 0 0 2 2
All 80 60 40 180
In [ ]:
sns.boxplot(data=df,y='Usage',x='Product')
plt.grid(axis='y')
In [ ]:
# we groupby the Education in years and product then count the frequency and plotted a scatterplot
plt.figure(figsize=(10,10))
k=sns.scatterplot(data=df.groupby(['Product','Usage']).size().reset_index().rename(columns={0:'values'}),
                  x='Usage',y='values',hue='Product')
k.set_xticks(range(1,8))
k.set_yticks(range(0,40,2))
plt.ylabel('No of Purchase')
plt.grid()
In [ ]:
# calculating the Conditional Probability 
Product_Usage_table['KP281|Usage']=round(Product_Usage_table['KP281']*100/Product_Usage_table['All'],2)
Product_Usage_table['KP481|Usage']=round(Product_Usage_table['KP481']*100/Product_Usage_table['All'],2)
Product_Usage_table['KP781|Usage']=round(Product_Usage_table['KP781']*100/Product_Usage_table['All'],2)
Product_Usage_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Usage KP481|Usage KP781|Usage
Usage
2 19 14 0 33 57.58 42.42 0.00
3 37 31 1 69 53.62 44.93 1.45
4 22 12 18 52 42.31 23.08 34.62
5 2 3 12 17 11.76 17.65 70.59
6 0 0 7 7 0.00 0.00 100.00
7 0 0 2 2 0.00 0.00 100.00
All 80 60 40 180 44.44 33.33 22.22
In [ ]:
# Plotting the countplot of Usages(per week) with the Product type
plt.figure(figsize=(15,10))
sns.countplot(x=df['Usage'],hue=df['Product'])
plt.title('Usages(per Week) VS Product',fontsize=20)
plt.show()

Observation-:

  1. KP281= more preferred for usage 2-4
  2. KP481= more preferred for usage 2-3
  3. KP781= more preferred for usage more than 4

Analyzation of Miles with Product¶

In [ ]:
df
Out[ ]:
Product Age Gender Education MaritalStatus Usage Fitness Income Miles Income (in K)
0 KP281 18 Male 14 Single 3 4 29562 112 29.56
1 KP281 19 Male 15 Single 2 3 31836 75 31.84
2 KP281 19 Female 14 Partnered 4 3 30699 66 30.70
3 KP281 19 Male 12 Single 3 3 32973 85 32.97
4 KP281 20 Male 13 Partnered 4 2 35247 47 35.25
... ... ... ... ... ... ... ... ... ... ...
175 KP781 40 Male 21 Single 6 5 83416 200 83.42
176 KP781 42 Male 18 Single 5 4 89641 200 89.64
177 KP781 45 Male 16 Single 5 5 90886 160 90.89
178 KP781 47 Male 18 Partnered 4 5 104581 120 104.58
179 KP781 48 Male 18 Partnered 4 5 95508 180 95.51

180 rows × 10 columns

In [ ]:
# Checking the outliers
plt.figure(figsize=(12,10))
g=sns.violinplot(data=df,y='Miles',x='Product')
g.set_yticks(range(0,400,25))
plt.title('Miles VS Product',fontsize=20)
plt.grid()
plt.show()
In [ ]:
plt.figure(figsize=(12,8))
h=sns.barplot(data=df,y='Miles',x='Product')
h.set_yticks(range(0,400,25))
plt.title('Miles VS Product',fontsize=20)
plt.grid()
plt.show()

OUtliers-

  • KP281 = miles more than 135 is outliers
  • KP481 = >125
  • KP781= >300
In [ ]:
#Extracting the Miles and product column and defining the Miles categories 
Product_Miles=df[['Product','Miles']]
a=[x for x in range(0,400,25)]
b=[str(a[i])+"-"+str(a[i+1]) for i in range(0,len(a)-1)]
Product_Miles['Miles_category']=pd.cut(Product_Miles['Miles'],bins=a,labels=b)
Product_Miles

C:\Users\Ravi Ahuja\AppData\Local\Temp\ipykernel_684\1078747982.py:5: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  Product_Miles['Miles_category']=pd.cut(Product_Miles['Miles'],bins=a,labels=b)
Out[ ]:
Product Miles Miles_category
0 KP281 112 100-125
1 KP281 75 50-75
2 KP281 66 50-75
3 KP281 85 75-100
4 KP281 47 25-50
... ... ... ...
175 KP781 200 175-200
176 KP781 200 175-200
177 KP781 160 150-175
178 KP781 120 100-125
179 KP781 180 175-200

180 rows × 3 columns

In [ ]:
product_miles_table=pd.crosstab(Product_Miles['Miles_category'],Product_Miles['Product'],margins=True)
product_miles_table
Out[ ]:
Product KP281 KP481 KP781 All
Miles_category
0-25 0 1 0 1
25-50 12 4 0 16
50-75 26 16 0 42
75-100 24 23 8 55
100-125 12 8 4 24
125-150 4 5 5 14
150-175 1 2 6 9
175-200 1 0 12 13
200-225 0 1 0 1
225-250 0 0 1 1
250-275 0 0 1 1
275-300 0 0 2 2
350-375 0 0 1 1
All 80 60 40 180
In [ ]:
# calculating the Conditional Probability 
product_miles_table['KP281|Miles']=round(product_miles_table['KP281']*100/product_miles_table['All'],2)
product_miles_table['KP481|Miles']=round(product_miles_table['KP481']*100/product_miles_table['All'],2)
product_miles_table['KP781|Miles']=round(product_miles_table['KP781']*100/product_miles_table['All'],2)
product_miles_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Miles KP481|Miles KP781|Miles
Miles_category
0-25 0 1 0 1 0.00 100.00 0.00
25-50 12 4 0 16 75.00 25.00 0.00
50-75 26 16 0 42 61.90 38.10 0.00
75-100 24 23 8 55 43.64 41.82 14.55
100-125 12 8 4 24 50.00 33.33 16.67
125-150 4 5 5 14 28.57 35.71 35.71
150-175 1 2 6 9 11.11 22.22 66.67
175-200 1 0 12 13 7.69 0.00 92.31
200-225 0 1 0 1 0.00 100.00 0.00
225-250 0 0 1 1 0.00 0.00 100.00
250-275 0 0 1 1 0.00 0.00 100.00
275-300 0 0 2 2 0.00 0.00 100.00
350-375 0 0 1 1 0.00 0.00 100.00
All 80 60 40 180 44.44 33.33 22.22
In [ ]:
df.describe()
Out[ ]:
Age Education Usage Fitness Income Miles Income (in K)
count 180.000000 180.000000 180.000000 180.000000 180.000000 180.000000 180.000000
mean 28.788889 15.572222 3.455556 3.311111 53719.577778 103.194444 53.720000
std 6.943498 1.617055 1.084797 0.958869 16506.684226 51.863605 16.507097
min 18.000000 12.000000 2.000000 1.000000 29562.000000 21.000000 29.560000
25% 24.000000 14.000000 3.000000 3.000000 44058.750000 66.000000 44.057500
50% 26.000000 16.000000 3.000000 3.000000 50596.500000 94.000000 50.595000
75% 33.000000 16.000000 4.000000 4.000000 58668.000000 114.750000 58.670000
max 50.000000 21.000000 7.000000 5.000000 104581.000000 360.000000 104.580000
In [ ]:
# we groupby the Miles and product then count the frequency and plotted a scatterplot
plt.figure(figsize=(10,10))
k=sns.scatterplot(data=df.groupby(['Product','Miles']).size().reset_index().rename(columns={0:'values'}),
                  x='Miles',y='values',hue='Product')
k.set_xticks(range(0,375,25))
k.set_yticks(range(0,18,1))
plt.ylabel('No of Purchase')
plt.grid()
plt.show()
In [ ]:
plt.figure(figsize=(15,10))
sns.countplot(data=Product_Miles,x='Miles_category',hue='Product')
plt.title('Miles_category VS Product',fontsize=20)
plt.show()
In [ ]:
# plotting the histogram for each product for Miles
fig=plt.figure(figsize=(20,8))
k=1
for i in ['KP281','KP481','KP781']:
    plt.subplot(1,3,k)
    sns.histplot(df.loc[df['Product']==i]['Miles'],bins=int(17),kde=True)
    plt.title(i)
    k=k+1
fig.suptitle('Checking the Miles for each Product Type',fontsize=25)
plt.show()

Observation -

  • KP 481 is preferred by customer expected to running miles in between(60-100)
  • less tha but it depends upon the income also KP281 is also purchased in between(80-100)
  • KP781 as it is an advance version so customer who are willing to running > 125 miles use this

Analyzation of Fitness with product¶

In [ ]:
product_fitness_table=pd.crosstab(df['Fitness'],df['Product'],margins=True)
product_fitness_table
Out[ ]:
Product KP281 KP481 KP781 All
Fitness
1 1 1 0 2
2 14 12 0 26
3 54 39 4 97
4 9 8 7 24
5 2 0 29 31
All 80 60 40 180
In [ ]:
# calculating the Conditional Probability 
product_fitness_table['KP281|Fitness']=round(product_fitness_table['KP281']*100/product_fitness_table['All'],2)
product_fitness_table['KP481|Fitness']=round(product_fitness_table['KP481']*100/product_fitness_table['All'],2)
product_fitness_table['KP781|Fitness']=round(product_fitness_table['KP781']*100/product_fitness_table['All'],2)
product_fitness_table
Out[ ]:
Product KP281 KP481 KP781 All KP281|Fitness KP481|Fitness KP781|Fitness
Fitness
1 1 1 0 2 50.00 50.00 0.00
2 14 12 0 26 53.85 46.15 0.00
3 54 39 4 97 55.67 40.21 4.12
4 9 8 7 24 37.50 33.33 29.17
5 2 0 29 31 6.45 0.00 93.55
All 80 60 40 180 44.44 33.33 22.22
In [ ]:
plt.figure(figsize=(15,10))
sns.countplot(data=df,x='Fitness',hue='Product')
plt.title('Miles_category VS Product',fontsize=20)
plt.show()

Observation:-

  1. customer having fitness score of 3 are tend to buy the KP281,KP481
  2. customer having fitness score of 5 are tend to buy the KP781.

CCN Analysis¶

In [ ]:
df
Out[ ]:
Product Age Gender Education MaritalStatus Usage Fitness Income Miles Income (in K)
0 KP281 18 Male 14 Single 3 4 29562 112 29.56
1 KP281 19 Male 15 Single 2 3 31836 75 31.84
2 KP281 19 Female 14 Partnered 4 3 30699 66 30.70
3 KP281 19 Male 12 Single 3 3 32973 85 32.97
4 KP281 20 Male 13 Partnered 4 2 35247 47 35.25
... ... ... ... ... ... ... ... ... ... ...
175 KP781 40 Male 21 Single 6 5 83416 200 83.42
176 KP781 42 Male 18 Single 5 4 89641 200 89.64
177 KP781 45 Male 16 Single 5 5 90886 160 90.89
178 KP781 47 Male 18 Partnered 4 5 104581 120 104.58
179 KP781 48 Male 18 Partnered 4 5 95508 180 95.51

180 rows × 10 columns

In [ ]:
male_df=df.loc[df['Gender']=="Male"]
male_df
Out[ ]:
Product Age Gender Education MaritalStatus Usage Fitness Income Miles Income (in K)
0 KP281 18 Male 14 Single 3 4 29562 112 29.56
1 KP281 19 Male 15 Single 2 3 31836 75 31.84
3 KP281 19 Male 12 Single 3 3 32973 85 32.97
4 KP281 20 Male 13 Partnered 4 2 35247 47 35.25
7 KP281 21 Male 13 Single 3 3 32973 85 32.97
... ... ... ... ... ... ... ... ... ... ...
175 KP781 40 Male 21 Single 6 5 83416 200 83.42
176 KP781 42 Male 18 Single 5 4 89641 200 89.64
177 KP781 45 Male 16 Single 5 5 90886 160 90.89
178 KP781 47 Male 18 Partnered 4 5 104581 120 104.58
179 KP781 48 Male 18 Partnered 4 5 95508 180 95.51

104 rows × 10 columns

In [ ]:
# Plotting the violin plot for Age with Product type and Gender category
plt.figure(figsize=(15,10))
q=sns.boxplot(data=df,y='Age',x='Gender',hue='Product')
q.set_yticks(range(0,55,2))
plt.title('Analysis of Age for males and female on each product type',fontsize=20)
plt.grid()
plt.show()
In [ ]:
# Plotting the violin plot for Age with Product type and Gender category
plt.figure(figsize=(15,10))
q=sns.violinplot(data=df,y='Age',x='Gender',hue='Product')
q.set_yticks(range(18,52,2))
plt.title('Analysis of Age for males and female on each product type',fontsize=20)
plt.grid(axis='y')
plt.show()

Observation-:

  1. Age 20-30:
    • Males= 24-26 has the higher chances for KP781
    • Females= 24-26 has the higher chances for KP781
  2. 30-35:
    • Males,Females = has the higher chances for KP481
  3. $>40$ there are higher chances of buying KP781,KP481 by males than females
  4. Median age of males buying KP481 is less than Median age of female buying KP481
In [ ]:
# Plotting the violin plot for Income with Product type and Gender category
plt.figure(figsize=(15,10))
q=sns.violinplot(data=df,y='Income',x='Gender',hue='Product')
q.set_yticks(range(0,130000,10000))
plt.title('Analysis of Income for males and female on each product type',fontsize=20)
plt.grid()
plt.show()
In [ ]:
# Plotting the violin plot for IUsage with Product type and Gender category
plt.figure(figsize=(15,10))
q=sns.violinplot(data=df,y='Usage',x='Gender',hue='Product')
q.set_yticks(range(0,7,1))
plt.title('Analysis of Usages for Female and Male on each product type',fontsize=20)
plt.grid()
plt.show()

Observation=

  1. There are more chances of Female having usage between 2-3 buying KP281,KP781 than males:
  2. Males and Females having higher income and more than 4 usage are tend to buy KP781
  3. Median use of Female is Higher than median use of males for KP781
In [ ]:
# Plotting the violin plot for Miles(Expected to run) with Product type and Gender category
plt.figure(figsize=(15,10))
q=sns.violinplot(data=df,y='Miles',x='Gender',hue='Product')
q.set_yticks(range(0,450,25))
plt.title('Analysis of Miles for Female and Male on each product type',fontsize=20)
plt.grid()
plt.show()

Observation:-

  1. Median of expected Miles running is higher than males using KP781
  2. For higher Miles use KP781
In [ ]:
# Plotting the violin plot for IFitness with Product type and Gender category
plt.figure(figsize=(15,10))
q=sns.violinplot(data=df,y='Fitness',x='Gender',hue='Product')
q.set_yticks(range(0,6))
plt.title('Analysis of Fitnessfor Female and Male on each product type',fontsize=20)
plt.grid()
plt.show()

Observation:-

  1. Usage=5 > KP781
  2. usages=3 > KP281,KP481
In [ ]:
# Plotting the violin plot for Education with Product type and Gender category
plt.figure(figsize=(15,10))
q=sns.violinplot(data=df,y='Education',x='Gender',hue='Product')
q.set_yticks(range(12,25))
plt.title('Analysis of Education for single and Male on each product type',fontsize=20)
plt.grid()
plt.show()
In [ ]:
# Plotting the violin plots of each Numerical column i.e. ['Age','Income','Miles','Education','Fitness','Usage'] 
# with Gender, Marital Status Status and Product Type

fig=plt.figure(figsize=(25,60))
k=1
for m in ['Gender','MaritalStatus']:
      for i in ['Age','Income','Miles','Education','Fitness','Usage']:
            plt.subplot(6,2,k)
            q=sns.violinplot(data=df,y=i,x=m,hue='Product')
            plt.title(f'Analysis of {i} for {m} on each product type',fontsize=20)
            plt.legend(loc='upper left',frameon=False,ncol=3)
            k=k+1
            if i=='Age':
                  q.set_yticks(range(0,50,5))
            elif i=='Income':
                  q.set_yticks(range(0,130000,10000))
            elif i=='Miles':
                  q.set_yticks(range(0,400,25))
            elif i=='Education':
                  q.set_yticks(range(10,26,2))
            plt.grid()
plt.show()
In [ ]:
# plotting the scatter plot for customer usages with customer income with each product type
plt.figure(figsize=(12,10))
t=sns.scatterplot(data=df,y='Usage',x='Income',hue='Product')
t.set_xticks(range(0,130000,5000))
plt.xticks(rotation=300)
plt.show()
In [ ]:
# plotting the scatter plot for customer Education with customer income with each product type
plt.figure(figsize=(12,10))
t=sns.scatterplot(data=df,y='Education',x='Income',hue='Product')
t.set_xticks(range(0,130000,5000))
t.set_yticks(range(10,24))
plt.xticks(rotation=300)
plt.show()
In [ ]:
# plotting the scatter plot for customer Expectation of miles running with customer income with each product type
plt.figure(figsize=(12,10))
t=sns.scatterplot(data=df,y='Miles',x='Income',hue='Product')
t.set_xticks(range(0,130000,5000))
t.set_yticks(range(0,425,25))
plt.xticks(rotation=300)
plt.show()
In [ ]:
# Plotting the scatter plot of all numerical column with each other for each product type 
fig=plt.figure(figsize=(30,40))
k=1
numlist=['Age','Income','Miles','Education','Fitness','Usage']
for i in range(0,6):
            for j in range(i+1,len(numlist)):
                  plt.subplot(5,3,k)
                  q=sns.scatterplot(data=df,x=numlist[i],y=numlist[j],hue='Product')
                  plt.title(f'Analysis of {numlist[i]} for {numlist[j]} on each product type',fontsize=18)
                  plt.legend(loc='upper left',frameon=False,ncol=3)
                  k=k+1
                  if (i) =='Age':
                        q.set_xticks(range(0,50,5))
                  plt.grid()
plt.show()
In [ ]:
# getting the correlation with each numerical with each other
df.corr()
Out[ ]:
Age Education Usage Fitness Income Miles Income (in K)
Age 1.000000 0.280496 0.015064 0.061105 0.513414 0.036618 0.513406
Education 0.280496 1.000000 0.395155 0.410581 0.625827 0.307284 0.625823
Usage 0.015064 0.395155 1.000000 0.668606 0.519537 0.759130 0.519548
Fitness 0.061105 0.410581 0.668606 1.000000 0.535005 0.785702 0.535007
Income 0.513414 0.625827 0.519537 0.535005 1.000000 0.543473 1.000000
Miles 0.036618 0.307284 0.759130 0.785702 0.543473 1.000000 0.543476
Income (in K) 0.513406 0.625823 0.519548 0.535007 1.000000 0.543476 1.000000
In [ ]:
# Plotting the heatmap 
sns.heatmap(df.corr(),cmap='coolwarm',annot=True)
Out[ ]:
<AxesSubplot:>
In [ ]:
# Plotting the pair plot 
sns.pairplot(df,hue='Product')
plt.show()
In [ ]:
# Plotting the Lineplot of all numerical column with each other for each product type 
fig=plt.figure(figsize=(20,40))
k=1
numlist=['Age','Income','Miles','Education','Fitness','Usage']
for i in range(0,6):
            for j in range(i+1,len(numlist)):
                  plt.subplot(5,3,k)
                  q=sns.lineplot(data=df,x=numlist[i],y=numlist[j],hue='Product',ci=False)
                  plt.title(f'Analysis of {numlist[i]} for {numlist[j]} on each product type',fontsize=12)
                  plt.legend(loc='upper left',frameon=False,ncol=3)
                  k=k+1
                  if (i) =='Age':
                        q.set_xticks(range(0,50,5))
                  plt.grid()
plt.show()

Observations:-

  1. Age-
    • KP281 - Customer having age in between 20-30
    • KP481 - Customer having age in between 30-35
    • KP781 - Customer having age in between >40(outliers),25-30
  2. Income-
    • KP281- customer having Income in between 30k-60k
    • KP481- customer having Income in between 40k-60k
    • KP781- customer having Income in greater > 60k
  3. Education-
    • KP281= customer having Education of 14 and 16 years
    • KP281= customer having Education of 14 and 16 years
    • KP781= customer having Education of 18 and more year
  4. Usages-
    • KP281= more preferred for usage 2-4
    • KP481= more preferred for usage 2-3
    • KP781= more preferred for usage more than 4
  5. Miles-
    • KP 481 is preferred by customer expected to running miles in between(60-100)
    • less than but it depends upon the income also KP281 is also purchased in between(80-100)
    • KP781 as it is an advance version so customer who are willing to running > 125 miles use this

  6. Fitness-

    • customer having fitness score of 3 are tend to buy the KP281,KP481
    • customer having fitness score of 5 are tend to buy the KP781.

  7. Age_Gender -

    • Age 20-30:
      • Males= 24-26 has the higher chances for KP781
      • Females= 24-26 has the higher chances for KP781
    • 30-35:
      • Males,Females = has the higher chances for KP481
    • $>40$ there are higher chances of buying KP781,KP481 by males than females
    • Median age of males buying KP481 is less than Median age of female buying KP481

  8. Usages-Gender:-

    • There are more chances of Female having usage between 2-3 buying KP281,KP781 than males:
    • Males and Females having higher income and more than 4 usage are tend to buy KP781
    • Median use of Female is Higher than median use of males for KP781

  9. Age_Income:-

    • KP281= Age(20-25), income(30k-50k)
    • KP481= Age(30-35), Income(45k-60k)
    • KP781 =Age(20-70)- Income(>60k)

  10. Age_Miles:-

    • KP281= Age(20-30), 50-100 Miles More preferred
    • KP481= Age(30-35), 50-100 Miles
    • KP781 =Age(25-30)- >125 miles
    • Age the age Increase people are tend to buy less KP781
  11. Age_Education:-
    • KP281= Age(20-25), Education(14)
    • KP481= Age(35-50), Education(16)
    • KP781= Age(25-30), Education (16)
    • KP781= Age(All), Education(18)
  12. Age_Fitness:-
    • KP281= Age(20-25), Fitness(3)
    • KP481= Age(35-50), Fitness(3)
    • KP781= Age(25-30), Fitness (5)
  13. Age_Usgaes:-
    • KP281= Age(20-25), Usage(2-3)
    • KP481= Age(35-50), Usage(2-3)
    • KP781= Age(5-30), Usage(>4)
  Observation-:
  1. KP281= more preferred for usage 2-4
  2. KP481= more preferred for usage 2-3
  3. KP781= more preferred for usage more than 4
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]: